The elucidation of molecular markers for class prediction and class discovery of disease states is an important area of medical research. Our goal is to develop arrays of protein markers for cancer diagnosis and prognosis using proteomics. Two-dimensional gel electrophoresis, mass spectrometry, and direct database searching programs have been combined to provide an enabling technology for proteomics. This initial enabling technology has allowed the full potential of proteomics to be envisioned, but limitations of the current analytical technologies used for the comprehensive analyses of proteins are hampering progress . The strength of the proteomic schemes for comprehensive analyses of cellular proteins is that no assumptions are made, making it possible to observe unanticipated interactions. Furthermore, such comprehensive methods provide a means to address the complexity of biological systems that is caused by the presence of many interdependent parallel signaling pathways. The objectives of this project are to: 1) over come these analytical limitations by developing an integrated analytical system capable of high throughput and high sensitivity protein analysis using mass spectrometry, and 2) apply these methods directly to model cell lines for the elucidation of cancer markers. The microfluidic sample handling and preparation device developed in this proposal will link multi-dimensional column separations with protein identification by mass spectrometry. This system will provide an attractive alternative to two-dimensional gel electrophoresis and in-gel digests. This system will allow the use of multi-dimensional column separations, which will greatly enhance the peak capacity of the protein separation making it possible to observe low abundance proteins in the presence of structural and housekeeping proteins. Furthermore, this system will reduce sample loss and contamination while decreasing the overall analysis time. Yeast cell lysates will be used to evaluate the performance of this system to detect low abundance proteins. Yeast were used to evaluate the current two-dimensional electrophoresis technology and will provide the most reliable comparison with this technique. For the elucidation of cancer markers we will begin with a well characterized system by comparing gene induction products of Src transform and non-transform mouse fibroblast cell lines. For further investigations we will analyze P13 kinase transformed cells and PI 3-kinase induced tumors from chick embryos for inducible protein expression. We anticipate that through the detection of low abundance proteins, we will be able to develop an array of protein markers that play an integral role in the mechanisms of cancer onset. Additionally, it is expected that in our search for regulatory markers we will gain new insights into the mechanisms of cell function and control.